<p>Applying power skiving to the machining of internal gears for new-energy electric motor transmissions can improve productivity and reduce manufacturing costs. However, its industrial application has been significantly hindered by the difficulty of achieving high tooth flank accuracy. This study focuses on the tooth flank morphology generated by power skiving and investigates the evolution of tooth flank ripples under the combined effects of multiple parameters. First, a theoretical model of the skived tooth flank was established based on the kinematic relationship of the skiving process, and an analytical method for deriving the normal deviation was developed, revealing the characteristics of its evolution. Subsequently, the influence weights of cutter and process parameters on the maximum normal deviation were quantitatively evaluated. Finally, based on the proposed tooth flank morphology model, experiments were conducted on a Y8030CNC gear skiving machine. The results show that the measured tooth flank is generally consistent with the simulated tooth flank, which verifies the validity of the proposed tooth flank morphology model under basic power skiving conditions. The main contribution of this study is the establishment of a tooth flank morphology model that explicitly describes the normal deviation distribution caused by discrete axial feed on the tooth flanks, providing a basis for tooth flank accuracy evaluation and parameter analysis.</p>